Block Term Decomposition Analysis in Long Segments of Atrial Fibrillation ECGs

Responsible for 25% of strokes, atrial fibrillation (AF) is the most common sustained cardiac arrhythmia in clinical practice. A precise analysis of the atrial activity (AA) signal in electrocardiogram (ECG) recordings is necessary to better understand this challenging cardiac condition. Recently, the block term decomposition (BTD) has been proposed as a novel tool to extract AA in AF ECG signals noninvasively. However, this tensor factorization technique was performed only in short segments. In this paper, the BTD is assessed in long segments of an AF ECG, varying the observation window size. Experimental results show the performance of BTD in long segments of an AF ECG recording and the analysis of the observation window size. This promising tensor technique is compared to two matrix-based methods.